Wikipedia:Requests for comment/Policies/Wikipedia:Requests for adminship
From Wikipedia, the free encyclopedia
You've all heard someone say "RfA is a vote" and then 10 angry users rebut it with "RfA is not a vote; the latter 10 are wrong, even though their stance abides by policy. RfA has a long history of users not being promoted due to the percentages in these two pages: User:Tangotango/RfA Analysis/Report and Wikipedia:Bureaucrats' noticeboard/RfA Report. Decisions on enwiki are determined by consensus not votes. On WP:BN, if someone is promoted with less than 75% or so, people start to rant how that is against policy. However, they — and likely the closing crats — don't even care about consensus if reading the table saves them the time from reading the discussion. RfA even has percentage recommendations for the closing bureaucrats. This infuriates me — and users for whom there was consensus to be promoted, but not the recommended support tally — to no end. Thank you. —« ANIMUM » 17:55, 28 July 2007 (UTC)
- Everyone is constantly saying how RFA is not a vote. But guess what? They're all lying, because RFA is a vote, with the big bold "support" and "oppose" comments, the numbering, the different sections, the silly vote counters that no one updates, and most of all, Tangobot's RFA report. If RFA is not a vote, why does it count the number of supports, opposes, and neutrals and gives the percentage of support? It does not help RFA in anyway, and it makes even more of a vote. The crats are supposed to determine consenus, not look at the little box and see it's green, so it must mean they're a good candidate, or look at the little box and see it's red, so it obviously means they're not a good candidate, and consenus has been reached of course because there's no way an RFA could be in the failing range, yet still have consenus. That is not what it means, but some crats think it is. The only useful feature of the report is that the bot tells you overdue nominations. That can be kept if needed, but the rest of the table should go. This is the first step in making RFA less of a vote, and making it so that all crats actually look at the RFA instead of just at this table. If RFA is not a vote, then you don't need a vote counter and percentage finder, do you? --(Review Me) R ParlateContribs@ (Let's Go Yankees!) 18:08, 28 July 2007 (UTC)
- It's not this simple. In AfD, all that matters is consensus that an article should be deleted/kept/whatever. Stupid rationales can be discounted easily, and strong arguments can be given more weight. But in RfA, it's different. What matters is whether or not someone would be a good admin, but a significant part of that is whether the community trusts them to be an admin. If the community doesn't trust them as much as we think should be required, it doesn't matter whether the lack of trust comes from 5 block log entries or lack of 5 featured articles. Admins can not use their powers well without trust. -Amarkov moo! 18:51, 28 July 2007 (UTC)
"This is the first step in making RFA less of a vote, and making it so that all crats actually look at the RFA instead of just at this table." – Just out of curiosity, do you have any evidence to support your claim that bureaucrats do not read RfA discussions as is their job and that they are actually so blatantly irresponsible as to base sysoppings solely on a bot-generated table? I respect you two very much, and this seems blatantly out of character. — Madman bum and angel (talk – desk) 18:56, 28 July 2007 (UTC)
- (ECx2) Is this really necessary? This topic has been beaten to death over and over again at WT:RFA and no agreement has ever resulted. Contary to Magnus animum's asseration at the beginning, RfA is not a vote (otherwise, we'd have a clear cutoff and no promotions below the 70-75% threshold). It is a mix of voting, discussion, and bureaucrat discretion. Unlike XfD, where consensus depends not on numbers but on arguments as they reflect policy, RfA is fundamentally about trust. It is not possible to legislate trust. As for the tables, leave them. Some people find them useful and others don't. I oppose placing them on the RfA talk page, but trying to get them deleted seems like an overreaction. Black Falcon (Talk) 18:57, 28 July 2007 (UTC)
- I agree wholeheartedly that RfA is based on trust, but there is not — has not, and will never be — a such thing as 70-75% trust. —« ANIMUM » 19:00, 28 July 2007 (UTC)
- True. RfA is certainly not a true vote, but it is also not like AfD where we can discount half the arguments if they're silly. -Amarkov moo! 19:02, 28 July 2007 (UTC)
- Why? I'm sure the crats disregard some people's votes. —« ANIMUM » 19:03, 28 July 2007 (UTC)
- What would be the definition of consensus in RfA then? Is it then different from consensus as in AfD, and how? - Mailer Diablo 19:04, 28 July 2007 (UTC)
- Why? I'm sure the crats disregard some people's votes. —« ANIMUM » 19:03, 28 July 2007 (UTC)
- Perhaps, but what is the alternative? Dictating to people whom or when to trust? Black Falcon (Talk) 19:07, 28 July 2007 (UTC)
- No, by people deciding per consensus if a user on RfA can be trusted. —« ANIMUM » 19:10, 28 July 2007 (UTC)
- And how does that happen? Sure, discussion is involved, but discussion already takes place to a certain extent (at least, there's plenty of opportunity for it on each RfA's talk page). But here's the key question: how do the crats decide whether there is consensus that a candidate can be trusted? If it involves discounting certain opinions (aside from trolls and WP:POINTs), then we're back to telling people what standards they should use to trust. Black Falcon (Talk) 19:15, 28 July 2007 (UTC)
- No, by people deciding per consensus if a user on RfA can be trusted. —« ANIMUM » 19:10, 28 July 2007 (UTC)
- True. RfA is certainly not a true vote, but it is also not like AfD where we can discount half the arguments if they're silly. -Amarkov moo! 19:02, 28 July 2007 (UTC)
- I agree wholeheartedly that RfA is based on trust, but there is not — has not, and will never be — a such thing as 70-75% trust. —« ANIMUM » 19:00, 28 July 2007 (UTC)
Has anyone actually established that RfA is broken? The impression is get is that RfA becomes irreconcilably broken when someone fails to suceed in their nomination, and is suddenly, magically, fixed when they succeed. Voting is not a bad thing, and can be significantly better than consensus in large discussions with a binary choice (which an RfA is). In most cases, counting by numbers is a good reflection of how the community feels - in others, more care may need to be taken to examine the opposers' reasons and discount the frivolous. Hell, I might even go so far as to say that allowing the community to express their opinion one of three ways might actually show the consensus. Nah - that could never work :\ Martinp23 19:06, 28 July 2007 (UTC)
-
- Martin, the situation that you seem to be imagining is that 75% support = candidate can be trusted and anything below he or she can't. The CharlotteWebb incident, for example, wherein Charlotte demonstrated she could be trusted, but everyone opposed for her use of Tor. There was the percentage to fail, but she clearly demonstrated that she could be trusted. —« ANIMUM » 19:20, 28 July 2007 (UTC)
- No - there is a discretionary range. It all depends on the opposes and how the 'crats view them. As it was, many people (exclusive of myself) felt that it was improper to have an admin breaking what was policy (since changed, I believe). Their opinions were just as valid as yours or mine. Martinp23 19:35, 28 July 2007 (UTC)
- Magnus animum, how was it clearly demonstrated that she could be trusted when 40% of participants said that they would not trust her? And keep in mind, please, that CharlotteWebb's RfA was one I particularly strongly supported. I thought and still think that the opposing reasons were ill-informed, but not everyone shares my or your standards for trusting candidates. Are RfA standards too high? In my opinion (and it's an opinion only), yes. The way to go about fixing that is to engage individuals who have high standards, rather than trying to have their opinions discounted. Black Falcon (Talk) 19:29, 28 July 2007 (UTC)
- My RFA failed with ~70% support. Does that mean I can't be trusted, and that as an admin I would go around deleting the main page and blocking Jimbo, or making bad decisions? --(Review Me) R ParlateContribs@ (Let's Go Yankees!) 19:23, 28 July 2007 (UTC)
- My opinion, given only what you just said, is that you can't be trusted as an admin, because that's the opinion that the 'crats were able to divine from your RfA. I have faith that, were the opposes frivolous (ie trolling), the 'crats would have promoted you. As it is, everyone is entitled to their opinion, and saying that we should go for some wispy concept called "consensus" just so that you can have a higher probability of passing RfA seems, to me, to be fatally flawed. Had you passed your recent RfA, would you be complaining now? Martinp23 19:35, 28 July 2007 (UTC)
- Yes I would. RFA is flawwed, and the process is too long and hard. Some users who would make amazing admins aren't because they don't feel like wasting time on RFA. Also, too many users are rejected because of edit count. Just because someone doesn't have x edits doesn't mean they're not good enough to be an admin. What if they're all major edits, where someone has x*10 edits than that person, but they're all minor? That needs to be fixed. --(Review Me) R ParlateContribs@ (Let's Go Yankees!) 19:47, 28 July 2007 (UTC)
- I don't disagree. But if we change the process to rely on "consensus" (in whatever way it can be applied here - IMO we've already got the best fit), those sorts of opinions won't go away, and will have to be taken into account by the closing 'crat regardless. What is needed is re-eduction. But then again, is it really that bad? I think it would be a lot worse if we went the way you suggest, allowing people to argue the oppositions of others out of significance. That's just going to cause more Wikihate. Martinp23 20:12, 28 July 2007 (UTC)
- Yes I would. RFA is flawwed, and the process is too long and hard. Some users who would make amazing admins aren't because they don't feel like wasting time on RFA. Also, too many users are rejected because of edit count. Just because someone doesn't have x edits doesn't mean they're not good enough to be an admin. What if they're all major edits, where someone has x*10 edits than that person, but they're all minor? That needs to be fixed. --(Review Me) R ParlateContribs@ (Let's Go Yankees!) 19:47, 28 July 2007 (UTC)
- My opinion, given only what you just said, is that you can't be trusted as an admin, because that's the opinion that the 'crats were able to divine from your RfA. I have faith that, were the opposes frivolous (ie trolling), the 'crats would have promoted you. As it is, everyone is entitled to their opinion, and saying that we should go for some wispy concept called "consensus" just so that you can have a higher probability of passing RfA seems, to me, to be fatally flawed. Had you passed your recent RfA, would you be complaining now? Martinp23 19:35, 28 July 2007 (UTC)
- Martin, the situation that you seem to be imagining is that 75% support = candidate can be trusted and anything below he or she can't. The CharlotteWebb incident, for example, wherein Charlotte demonstrated she could be trusted, but everyone opposed for her use of Tor. There was the percentage to fail, but she clearly demonstrated that she could be trusted. —« ANIMUM » 19:20, 28 July 2007 (UTC)
- I have written an essay on the flaws of RFA. --(Review Me) R ParlateContribs@ (Let's Go Yankees!) 19:13, 28 July 2007 (UTC)
- We know RFA is corrupt. The process needs review. Complaining (which, I'm sorry R, seems to be what your essay is for) will do no good. The flaws of RFA are evident enough; we dont need an essay to identify them for us. Lets work on finding solutions, new ways of doing things, instead of constantly remarking, complaining, commenting on and policy slapping each other at WT:RFA and else where. Instead of roiling around, constantly debating what is wrong with RFA, lets debate what can be done about it; what we should do about it. How about it poeple? -- Anonymous DissidentTalk 06:49, 29 July 2007 (UTC)
[edit] Resolution
The core issue at hand seems to be that while the policy says the RfA is a consensus-driven process, the reality is different. As a result, there's an eternal struggle between those who point out that "flaw" and those who uphold the consensus idea, even if it's not true. Therefore, I hereby declare what follows:
- RfA is a vote.
Ok, now that it's openly admitted, can we return to building the encyclopedia, please? —Миша13 21:59, 28 July 2007 (UTC) (PS: Someone care to update the WP:RfA page to reflect this resolution? Thanks.)
Let's not confuse the potential flaws of RfA with every editor's frustration at seeing his own or some candidate's RfA failure. No system will ever avoid both false positive and false negatives. Tangobot's report or the vote tallies are not to blame for CharlotteWeb's failed RfA: it failed because of the widespread unease with open proxies. Whether or not that unease is a sound attitude from the community is an entirely different debate. More often than not, bureaucrats have taken the time to explain how they determined consensus in borderline cases and there is no evidence that they "don't even care about consensus if reading the table saves them the time from reading the discussion." Pascal.Tesson 22:23, 28 July 2007 (UTC)
- By the CharlotteWebb example, it was to illustrate that Charlotte could be trusted, not whether Tangobot was to blame. —« ANIMUM » 01:04, 29 July 2007 (UTC)
- Clearly some people (a fair number) felt that she couldn't be. We can't ignore their feelings, even if we disagree with them. This is why this is one of the most flawed RfA reform proposals I've seen (though, to be honest, it's not too different from many): "We need to change RfA so that the 'crats can come to a decision which supports my viewpoint". Grr - get back to building an encyclopedia. Martinp23 01:19, 29 July 2007 (UTC)
